arxiv:2605.17887
Haoran Dai
haorandai
AI & ML interests
None yet
Recent Activity
authored a paper 14 days ago
Attention Sinks and Outliers in Attention Residuals upvoted a paper 14 days ago
Attention Sinks and Outliers in Attention Residuals upvoted a paper 3 months ago
On-Policy Self-Distillation without Any SupervisionOrganizations
None yet
models 42
haorandai/deepseek-r1-llama-8b-strongest-vector
8B • Updated • 7
haorandai/qwen3-8b-gold-steered-teacher-500
8B • Updated • 1
haorandai/qwen3-8b-gold-steered-teacher
8B • Updated • 4
haorandai/qwen3-8b-gold-60k-alpha50
Text Generation • 8B • Updated • 4
haorandai/r1-qwen-7b-gold-60k-alpha50
Text Generation • 8B • Updated • 6
haorandai/qwen3-8b-gold-moredata
Text Generation • 8B • Updated • 3
haorandai/r1-qwen-7b-gold-moredata
Text Generation • 8B • Updated • 7
haorandai/r1-llama-8b-gold-moredata
Text Generation • 8B • Updated • 5
haorandai/qwen3-8b-gold-openr1
Text Generation • 8B • Updated • 5
haorandai/deepseek-r1-qwen-7b-gold
Text Generation • 8B • Updated • 20
datasets 117
haorandai/test_time_GSM8K_OpenR1
Updated • 3
haorandai/openr1-math-1k-subset-correct
Updated • 4 • 1
haorandai/JailbreakDB
Updated • 16 • 1
haorandai/AdaptiveDefense_Jan14_PGD_Mice_UF_epsilon0.1_10samples_3constraints
Viewer • Updated • 13 • 2
haorandai/AdaptiveDefense_Jan14_PGD_Bicycle_UF_epsilon0.1_10samples_3constraints
Viewer • Updated • 13 • 10
haorandai/AdaptiveDefense_Jan14_PGD_Banana_UF_epsilon0.1_10samples_3constraints
Viewer • Updated • 13 • 3
haorandai/AdaptiveDefense_Feb10_PGD_Mice_Chessboard_epsilon0.1_10samples_3constraints
Viewer • Updated • 13 • 7
haorandai/AdaptiveDefense_Feb10_PGD_Bicycle_Chessboard_epsilon0.1_10samples_3constraints
Viewer • Updated • 13 • 6
haorandai/AdaptiveDefense_Feb10_PGD_Banana_Chessboard_epsilon0.1_10samples_3constraints
Viewer • Updated • 13 • 6
haorandai/Defense_Jan14_PGD_Mice_Orange_epsilon0.1_10samples_3constraints
Viewer • Updated • 13 • 3